Positive Feedbacks among Forest Fragmentation, Drought, and Climate Change in the Amazon

Abstract: The Amazon basin is experiencing rapid forest loss and fragmentation. Fragmented forests are more prone than intact forests to periodic damage from El Niño–Southern Oscillation ( ENSO) droughts, which cause elevated tree mortality, increased litterfall, shifts in plant phenology, and other ecological changes, especially near forest edges. Moreover, positive feedbacks among forest loss, fragmentation, fire, and regional climate change appear increasingly likely. Deforestation reduces plant evapotranspiration, which in turn constrains regional rainfall, increasing the vulnerability of forests to fire. Forest fragments are especially vulnerable because they have dry, fire-prone edges, are logged frequently, and often are adjoined by cattle pastures, which are burned regularly. The net result is that there may be a critical "deforestation threshold" above which Amazonian rainforests can no longer be sustained, particularly in relatively seasonal areas of the basin. Global warming could exacerbate this problem if it promotes drier climates or stronger ENSO droughts. Synergisms among many simultaneous environmental changes are posing unprecedented threats to Amazonian forests.     

Climate effects on fire regimes and tree recruitment in Black Hills ponderosa pine forests

Climate influences forest structure through effects on both species demography (recruitment and mortality) and disturbance regimes. Here, I compare multi-century chronologies of regional fire years and tree recruitment from ponderosa pine forests in the Black Hills of southwestern South Dakota and northeastern Wyoming to reconstructions of precipitation and global circulation indices. Regional fire years were affected by droughts and variations in both Pacific and Atlantic sea surface temperatures. Fires were synchronous with La Ni?as, cool phases of the Pacific Decadal Oscillation (PDO), and warm phases of the Atlantic Multidecadal Oscillation (AMO). These quasi-periodic circulation features are associated with drought conditions over much of the western United States. The opposite pattern (El Ni?o, warm PDO, cool AMO) was associated with fewer fires than expected. Regional tree recruitment largely occurred during wet periods in precipitation reconstructions, with the most abundant recruitment coeval with an extended pluvial from the late 1700s to early 1800s. Widespread even-aged cohorts likely were not the result of large crown fires causing overstory mortality, but rather were caused by optimal climate conditions that contributed to synchronous regional recruitment and longer intervals between surface fires. Synchronous recruitment driven by climate is an example of the Moran effect. The presence of abundant fire-scarred trees in multi-aged stands supports a prevailing historical model for ponderosa pine forests in which recurrent surface fires affected heterogenous forest structure, although the Black Hills apparently had a greater range of fire behavior and resulting forest structure over multi-decadal time scales than ponderosa pine forests of the Southwest that burned more often.     

Mortality of large trees and lianas following experimental drought in an Amazon forest

Severe drought episodes such as those associated with El Ni?o Southern Oscillation (ENSO) events influence large areas of tropical forest and may become more frequent in the future. One of the most important forest responses to severe drought is tree mortality, which alters forest structure, composition, carbon content, and flammability, and which varies widely. This study tests the hypothesis that tree mortality increases abruptly during drought episodes when plant-available soil water (PAW) declines below a critical minimum threshold. It also examines the effect of tree size, plant life form (palm, liana, tree) and potential canopy position (understory, midcanopy, overstory) on drought-induced plant mortality. A severe, four-year drought episode was simulated by excluding 60% of incoming throughfall during each wet season using plastic panels installed in the understory of a 1-ha forest treatment plot, while a 1-ha control plot received normal rainfall. After 3.2 years, the treatment resulted in a 38% increase in mortality rates across all stems >2 cm dbh. Mortality rates increased 4.5-fold among large trees (>30 cm dbh) and twofold among medium trees (10-30 cm dbh) in response to the treatment, whereas the smallest stems were less responsive. Recruitment rates did not compensate for the elevated mortality of larger-diameter stems in the treatment plot. Overall, lianas proved more susceptible to drought-induced mortality than trees or palms, and potential overstory tree species were more vulnerable than midcanopy and understory species. Large stems contributed to 90% of the pretreatment live aboveground biomass in both plots. Large-tree mortality resulting from the treatment generated 3.4 times more dead biomass than the control plot. The dramatic mortality response suggests significant, adverse impacts on the global carbon cycle if climatic changes follow current trends.     

Life history trade-offs in tropical trees and lianas

It has been hypothesized that tropical trees partition forest light environments through a life history trade-off between juvenile growth and survival; however, the generality of this trade-off across life stages and functional groups has been questioned. We quantified trade-offs between growth and survival for trees and lianas on Barro Colorado Island (BCI), Panama using first-year seedlings of 22 liana and 31 tree species and saplings (10 mm < dbh < 39 mm) of 30 tree species. Lianas showed trade-offs similar to those of trees, with both groups exhibiting broadly overlapping ranges in survival and relative growth rates as seedlings. Life history strategies at the seedling stage were highly correlated with those at the sapling stage among tree species, with all species showing an increase in survival with size. Only one of 30 tree species demonstrated a statistically significant ontogenetic shift, having a relatively lower survival rate at the sapling stage than expected. Our results indicate that similar life history trade-offs apply across two functional groups (lianas and trees), and that life history strategies are largely conserved across seedling and sapling life-stages for most tropical tree species.     

Effects of Pioneer Tree Species Hyperabundance on Forest Fragments in Northeastern Brazil

Abstract: Despite many studies on fragmentation of tropical forests, the extent to which plant and animal communities are altered in small, isolated forest fragments remains obscure if not controversial. We examined the hypothesis that fragmentation alters the relative abundance of tree species with different vegetative and reproductive traits. In a fragmented landscape (670 km²) of the Atlantic Forest of northeastern Brazil, we categorized 4056 trees of 182 species by leafing pattern, reproductive phenology, and morphology of seeds and fruit. We calculated relative abundance of traits in 50 1‐ha plots in three types of forest configurations: forest edges, small forest fragments (3.4–83.6 ha), and interior of the largest forest fragment (3500 ha, old growth). Although evergreen species were the most abundant across all configurations, forest edges and small fragments had more deciduous and semideciduous species than interior forest. Edges lacked supra‐annual flowering and fruiting species and had more species and stems with drupes and small seeds than small forest fragments and forest interior areas. In an ordination of species similarity and life‐history traits, the three types of configurations formed clearly segregated clusters. Furthermore, the differences in the taxonomic and functional (i.e., trait‐based) composition of tree assemblages we documented were driven primarily by the higher abundance of pioneer species in the forest edge and small forest fragments. Our work provides strong evidence that long‐term transitions in phenology and seed and fruit morphology of tree functional groups are occurring in fragmented tropical forests. Our results also suggest that edge‐induced shifts in tree assemblages of tropical forests can be larger than previously documented.     

Woody plant richness and NDVI response to drought events in Catalonian (northeastern Spain) forests

The role of species diversity on ecosystem resistance in the face of strong environmental fluctuations has been addressed from both theoretical and experimental viewpoints to reveal a variety of positive and negative relationships. Here we explore empirically the relationship between the richness of forest woody species and canopy resistance to extreme drought episodes. We compare richness data from an extensive forest inventory to a temporal series of satellite imagery that estimated drought impact on forest canopy as NDVI (normalized difference vegetation index) anomalies of the dry summer in 2003 in relation to records of previous years. We considered five different types of forests that are representative of the main climatic and altitudinal gradients of the region, ranging from lowland Mediterranean to mountain boreal-temperate climates. The observed relationship differed among forest types and interacted with the climate, summarised by the Thorntwaite index. In Mediterranean Pinus halepensis forests, NDVI decreased during the drought. This decrease was stronger in forests with lower richness. In Mediterranean evergreen forests of Quercus ilex, drought did not result in an overall NDVI loss, but lower NDVI values were observed in drier localities with lower richness, and in more moist localities with higher number of species. In mountain Pinus sylvestris forests NDVI decreased, mostly due to the drought impact on drier localities, while no relation to species richness was observed. In moist Fagus sylvatica forests, NDVI only decreased in plots with high richness. No effect of drought was observed in the high mountain Pinus uncinata forests. Our results show that a shift on the diversity-stability relationship appears across the regional, climatic gradient. A positive relationship appears in drier localities, supporting a null model where the probability of finding a species able to cope with drier conditions increases with the number of species. However, in more moist localities we hypothesize that the proportion of drought-sensitive species would increase in richer localities, due to a higher likelihood of co-occurrence of species that share moist climatic requirements. The study points to the convenience of considering the causes of disturbance in relation to current environmental gradients and historical environmental constraints on the community.     

Projected vegetation changes for the American Southwest: combined dynamic modeling and bioclimatic-envelope approach

This study focuses on potential impacts of 21st century climate change on vegetation in the Southwest United States, based on debiased and interpolated climate projections from 17 global climate models used in the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Among these models a warming trend is universal, but projected changes in precipitation vary in sign and magnitude. Two independent methods are applied: a dynamic global vegetation model to assess changes in plant functional types and bioclimatic envelope modeling to assess changes in individual tree and shrub species and biodiversity. The former approach investigates broad responses of plant functional types to climate change, while considering competition, disturbances, and carbon fertilization, while the latter approach focuses on the response of individual plant species, and net biodiversity, to climate change. The dynamic model simulates a region-wide reduction in vegetation cover during the 21st century, with a partial replacement of evergreen trees with grasses in the mountains of Colorado and Utah, except at the highest elevations, where tree cover increases. Across southern Arizona, central New Mexico, and eastern Colorado, grass cover declines, in some cases abruptly. Due to the prevalent warming trend among all 17 climate models, vegetation cover declines in the 21st century, with the greatest vegetation losses associated with models that project a drying trend. The inclusion of the carbon fertilization effect largely ameliorates the projected vegetation loss. Based on bioclimatic envelope modeling for the 21st century, the number of tree and shrub species that are expected to experience robust declines in range likely outweighs the number of species that are expected to expand in range. Dramatic shifts in plant species richness are projected, with declines in the high-elevation evergreen forests, increases in the eastern New Mexico prairies, and a northward shift of the Sonoran Desert biodiversity maximum.     

Neutral theory and the evolution of ecological equivalence

Since the publication of the unified neutral theory in 2001, there has been much discussion of the theory, pro and con. The hypothesis of ecological equivalence is the fundamental yet controversial idea behind neutral theory. Assuming trophically similar species are demographically alike (symmetric) on a per capita basis is only an approximation, but it is equivalent to asking: How many of the patterns of ecological communities are the result of species similarities, rather than of species differences? The strategy behind neutral theory is to see how far one can get with the simplification of assuming ecological equivalence before introducing more complexity. In another paper, I review the empirical evidence that led me to hypothesize ecological equivalence among many of the tree species in the species-rich tropical forest on Barro Colorado Island (BCI). In this paper, I develop a simple model for the evolution of ecological equivalence or niche convergence, using as an example evolution of the suite of life history traits characteristic of shade tolerant tropical tree species. Although the model is simple, the conclusions from it seem likely to be robust. I conclude that ecological equivalence for resource use are likely to evolve easily and often, especially in species-rich communities that are dispersal and recruitment limited. In the case of the BCI forest, tree species are strongly dispersal- and recruitment-limited, not only because of restricted seed dispersal, but also because of low recruitment success due to heavy losses of the seedling stages to predators and pathogens and other abiotic stresses such as drought. These factors and the high species richness of the community strongly reduce the potential for competitive exclusion of functionally equivalent or nearly equivalent species.     

A tree and climate assessment tool for modelling ecosystem response to climate change

Understanding how vulnerable forest ecosystems are to climate change is a key requirement if sustainable forest management is to be achieved. Modelling the response of species in their regeneration niche to phenological and biophysical processes that are directly influenced by climate is one method for achieving this understanding. A model was developed to investigate species resilience and vulnerability to climate change within its fundamental-regeneration niche. The utility of the developed model, tree and climate assessment (TACA), was tested within the interior Douglas-fir ecosystem in south-central British Columbia. TACA modelled the current potential tree species composition of the ecosystem with high accuracy and modelled significant responses amongst tree species to climate change. The response of individual species suggests that the studied ecosystem could transition to a new ecosystem over the next 100 years. TACA showed that it can be an effective tool for identifying species resilience and vulnerability to changes in climate within the most sensitive stage of development, the regeneration phase. The TACA model was able to identify the degree of change in phenological and biophysical variables that control tree establishment, growth and persistence. The response to changes in one or more of these variables resulted in changes in the climatic suitability of the ecosystem for species and enabled a measure of vulnerability to be quantified. TACA could be useful to forest managers as a decision support tool for adaptation actions and by researchers interested in modelling stand dynamics under climate change.     

Changes in microbial community characteristics and soil organic matter with nitrogen additions in two tropical forests

Microbial communities and their associated enzyme activities affect the amount and chemical quality of carbon (C) in soils. Increasing nitrogen (N) deposition, particularly in N-rich tropical forests, is likely to change the composition and behavior of microbial communities and feed back on ecosystem structure and function. This study presents a novel assessment of mechanistic links between microbial responses to N deposition and shifts in soil organic matter (SOM) quality and quantity. We used phospholipid fatty acid (PLFA) analysis and microbial enzyme assays in soils to assess microbial community responses to long-term N additions in two distinct tropical rain forests. We used soil density fractionation and 13C nuclear magnetic resonance (NMR) spectroscopy to measure related changes in SOM pool sizes and chemical quality. Microbial biomass increased in response to N fertilization in both tropical forests and corresponded to declines in pools of low-density SOM. The chemical quality of this soil C pool reflected ecosystem-specific changes in microbial community composition. In the lower-elevation forest, there was an increase in gram-negative bacteria PLFA biomass, and there were significant losses of labile C chemical groups (O-alkyls). In contrast, the upper-elevation tropical forest had an increase in fungal PLFAs with N additions and declines in C groups associated with increased soil C storage (alkyls). The dynamics of microbial enzymatic activities with N addition provided a functional link between changes in microbial community structure and SOM chemistry. Ecosystem-specific changes in microbial community composition are likely to have far-reaching effects on soil carbon storage and cycling. This study indicates that microbial communities in N-rich tropical forests can be sensitive to added N, but we can expect significant variability in how ecosystem structure and function respond to N deposition among tropical forest types.     

Assessing the strength of climate and land-use influences on montane epiphyte communities

Epiphytes, air plants that are structurally dependent on trees, are a keystone group in tropical forests; they support the food and habitat needs of animals and influence water and nutrient cycles. They reach peak diversity in humid montane forests. Climate predictions for Central American mountains include increased temperatures, altered precipitation seasonality, and increased cloud base heights, all of which may challenge epiphytes. Although remaining montane forests are highly fragmented, many tropical agricultural systems include trees that host epiphytes, allowing epiphyte communities to persist even in landscapes with lower forest connectivity. I used structural equations models to test the relative effects of climate, land use, tree characteristics, and biotic interactions on vascular epiphyte diversity with data from 31 shade coffee farms and 2 protected forests in northern Nicaragua. I also tested substrate preferences of common species with randomization tests. Tree size, tree diversity, and climate all affected epiphyte richness, but the effect of climate was almost entirely mediated by bryophyte cover. Bryophytes showed strong sensitivity to mean annual temperature and insolation. Many ferns and some orchids were positively associated with bryophyte mats, whereas bromeliads tended to establish among lichen or on bare bark. The tight relationships between bryophytes and climate and between bryophytes and vascular epiphytes indicated that relatively small climate changes could result in rapid, cascading losses of montane epiphyte communities. Currently, shade coffee farms can support high bryophyte cover and diverse vascular epiphyte assemblages when larger, older trees are present. Agroforests serve as valuable reservoirs for epiphyte biodiversity and may be important early-warning systems as the climate changes.     

Avian Survival Rates and the Extinction Process on Barro Colorado Island, Panama

Abstract: Selective extinction following isolation of habitat patches may be due to biogeographical (e.g., island size or isolation) and ecological (species natural histories, interspecifc interactions) factors, or their interactions. Among the demographic and life history attributes commonly associated with high extinction probability are small populations, large size of individuals, and population variability. Long-term capture-recapture data from forest habitat in central Panama permit an examination of the association between mainland survival rates and extinction on a nearby land-bridge island Species of birds that no longer occur on Barro Colorado Island (BCI), Panama, have, on average, lower survival rates on the adjacent mainland than species that have persisted on BCI. Moreover, of the species that no longer occur on BCI, those with lower mainland survival rates generally disappeared earlier from the island. My analysis provides little evidence of a relationship between extinction and population size. Recolonization of BCI from the adjacent mainland by the forest undergrowth species studied here is unlikely. Reduced reproductive success on BCI combined with naturally low adult survival rates seems to be responsible for these BCI extinctions. High nest predation and/or altered landscape dynamics are probable agents in the low reproductive success. The methods used here could be employed in other circumstances to identify fragmentation-sensitive species.     

Common ectomycorrhizal networks may maintain monodominance in a tropical rain forest

Most tropical rain forests contain diverse arrays of tree species that form arbuscular mycorrhizae. In contrast, the less common monodominant rain forests, in which one tree species comprises more than 50% of the canopy, frequently contain ectomycorrhizal (ECM) associates. In this study, I explored the potential for common ECM networks, created by aggregations of ECM trees, to enhance seedling survivorship near parent trees. I determined the benefit conferred by the common ECM network on seedling growth and survivorship of an ECM monodominant species in Guyana. Seedlings with access to an ECM network had greater growth (73% greater), leaf number (55% more), and survivorship (47% greater) than seedlings without such access, suggesting that the ECM network provides a survivorship advantage. A survey of wild seedlings showed positive distance-dependent distribution and survival with respect to conspecific adults. These experimental and survey results suggest that the negative distance-dependent mechanisms at the seedling stage thought to maintain tropical rain forest diversity are reversed for ECM seedlings, which experience positive feedbacks from the ECM network. These results may in part explain the local monodominance of an ECM tree species within the matrix of high-diversity, tropical rain forest.     

Beyond Kyoto: Forest Management in a Time of Rapid Climate Change

Abstract: Policies to reduce global warming by offering credits for carbon sequestration have neglected the effects of forest management on biodiversity. I review properties of forest ecosystems and management options for enhancing the resistance and resilience of forests to climate change. Although forests, as a class, have proved resilient to past changes in climate, today's fragmented and degraded forests are more vulnerable. Adaptation of species to climate change can occur through phenotypic plasticity, evolution, or migration to suitable sites, with the latter probably the most common response in the past. Among the land-use and management practices likely to maintain forest biodiversity and ecological functions during climate change are (1) representing forest types across environmental gradients in reserves; (2) protecting climatic refugia at multiple scales; (3) protecting primary forests; (4) avoiding fragmentation and providing connectivity, especially parallel to climatic gradients; (5) providing buffer zones for adjustment of reserve boundaries; (6) practicing low-intensity forestry and preventing conversion of natural forests to plantations; ( 7) maintaining natural fire regimes; (8) maintaining diverse gene pools; and (9) identifying and protecting functional groups and keystone species. Good forest management in a time of rapidly changing climate differs little from good forest management under more static conditions, but there is increased emphasis on protecting climatic refugia and providing connectivity.     

A climatic driver for abrupt mid-Holocene vegetation dynamics and the hemlock decline in New England

The mid-Holocene decline of eastern hemlock is widely viewed as the sole prehistorical example of an insect- or pathogen-mediated collapse of a North American tree species and has been extensively studied for insights into pest-host dynamics and the consequences to terrestrial and aquatic ecosystems of dominant-species removal. We report paleoecological evidence implicating climate as a major driver of this episode. Data drawn from sites across a gradient in hemlock abundance from dominant to absent demonstrate: a synchronous, dramatic decline in a contrasting taxon (oak); changes in lake sediments and aquatic taxa indicating low water levels; and one or more intervals of intense drought at regional to continental scales. These results, which accord well with emerging climate reconstructions, challenge the interpretation of a biotically driven hemlock decline and highlight the potential for climate change to generate major, abrupt dynamics in forest ecosystems.     

Bird extirpations and community dynamics in an Andean cloud forest over 100 years of land-use change

Long-term studies to understand biodiversity changes remain scarce—especially so for tropical mountains. We examined changes from 1911 to 2016 in the bird community of the cloud forest of San Antonio, a mountain ridge in the Colombian Andes. We evaluated the effects of past land-use change and assessed species vulnerability to climate disruption. Forest cover decreased from 95% to 50% by 1959, and 33 forest species were extirpated. From 1959 to 1990, forest cover remained stable, and an additional 15 species were lost—a total of 29% of the forest bird community. Thereafter, forest cover increased by 26% and 17 species recolonized the area. The main cause of extirpations was the loss of connections to adjacent forests. Of the 31 (19%) extirpated birds, 25 have ranges peripheral to San Antonio, mostly in the lowlands. Most still occurred regionally, but broken forest connections limited their recolonization. Other causes of extirpation were hunting, wildlife trade, and water diversion. Bird community changes included a shift from predominantly common species to rare species; forest generalists replaced forest specialists that require old growth, and functional groups, such as large-body frugivores and nectarivores, declined disproportionally. All water-dependent birds were extirpated. Of the remaining 122 forest species, 19 are vulnerable to climate disruption, 10 have declined in abundance, and 4 are threatened. Our results show unequivocal species losses and changes in community structure and abundance at the local scale. We found species were extirpated after habitat loss and fragmentation, but forest recovery stopped extirpations and helped species repopulate. Land-use changes increased species vulnerability to climate change, and we suggest reversing landscape transformation may restore biodiversity and improve resistance to future threats.     

Climate determines upper, but not lower, altitudinal range limits of Pacific Northwest conifers

Does climate determine species' ranges? Rapid rates of anthropogenic warming make this classic ecological question especially relevant. We ask whether climate controls range limits by quantifying relationships between climatic variables (precipitation, temperature) and tree growth across the altitudinal ranges of six Pacific Northwestern conifers on Mt. Rainier, Washington, USA. Results for three species (Abies amabilis, Callitropsis nootkatensis, Tsuga mertensiana) whose upper limits occur at treeline (> 1600 m) imply climatic controls on upper range limits, with low growth in cold and high snowpack years. Annual growth was synchronized among individuals at upper limits for these high-elevation species, further suggesting that stand-level effects such as climate constrain growth more strongly than local processes. By contrast, at lower limits climatic effects on growth were weak for these high-elevation species. Growth-climate relationships for three low-elevation species (Pseudotsuga menziesii, Thuja plicata, Tsuga heterophylla) were not consistent with expectations of climatic controls on upper limits, which are located within closed-canopy forest (< 1200 m). Annual growth of these species was poorly synchronized among individuals. Our results suggest that climate controls altitudinal range limits at treeline, while local drivers (perhaps biotic interactions) influence growth in closed-canopy forests. Climate-change-induced range shifts in closed-canopy forests will therefore be difficult to predict accurately.     

Climate change and the cost of conserving species in Madagascar

We examined the cost of conserving species as climate changes. We used a Maxent species distribution model to predict the ranges from 2000 to 2080 of 74 plant species endemic to the forests of Madagascar under 3 climate scenarios. We set a conservation target of achieving 10,000 ha of forest cover for each species and calculated the cost of achieving this target under each scenario. We interviewed managers of projects to restore native forests and conducted a literature review to obtain the net present cost per hectare of management actions to maintain or establish forest cover. For each species, we added hectares of land from lowest to highest cost per additional year of forest cover until the conservation target was achieved throughout the time period. Climate change was predicted to reduce the size of species' ranges, the overlap between species' ranges and existing or planned protected areas, and the overlap between species' ranges and existing forest. As a result, climate change increased the cost of achieving the conservation target by necessitating successively more costly management actions: additional management within existing protected areas (US$0-60/ha); avoidance of forest degradation (i.e., loss of biomass) in community-managed areas ($160-576/ha); avoidance of deforestation in unprotected areas ($252-1069/ha); and establishment of forest on nonforested land within protected areas ($802-2710/ha), in community-managed areas ($962-3226/ha), and in unprotected areas ($1054-3719/ha). Our results suggest that although forest restoration may be required for the conservation of some species as climate changes, it is more cost-effective to maintain existing forest wherever possible.     

Seasonal water stress tolerance and habitat associations within four neotropical tree genera

We investigated the relationship between habitat association and physiological performance in four congeneric species pairs exhibiting contrasting distributions between seasonally flooded and terra firme habitats in lowland tropical rain forests of French Guiana, including Virola and Iryanthera (Myristicaceae), Symphonia (Clusiaceae), and Eperua (Caesalpiniaceae). We analyzed 10-year data sets of mapped and measured saplings (stems >150 cm in height and <10 cm diameter at breast height [dbh]) and trees (stems > or =10 cm dbh) across 37.5 ha of permanent plots covering a 300-ha zone, within which seasonally flooded areas (where the water table never descends below 1 m) have been mapped. Additionally, we tested the response of growth, survival, and leaf functional traits of these species to drought and flood stress in a controlled experiment. We tested for habitat preference using a modification of the torus translation method. Strong contrasting associations of the species pairs of Iryanthera, Virola, and Symphonia were observed at the sapling stage, and these associations strengthened for the tree stage. Neither species of Eperua was significantly associated with flooded habitats at the sapling stage, but E. falcata was significantly and positively associated with flooded forests at the tree stage, and trees of E. grandiflora were found almost exclusively in nonflooded habitats. Differential performance provided limited explanatory support for the observed habitat associations, with only congeners of Iryanthera exhibiting divergent sapling survival and tree growth. Seedlings of species associated with flooded forest tended to have higher photosynthetic capacity than their congeners at field capacity. In addition, they tended to have the largest reductions in leaf gas exchange and growth rate in response to experimental drought stress and the least reductions in response to experimental inundation. The corroboration of habitat association with differences in functional traits and, to a lesser extent, measures of performance provides an explanation for the regional coexistence of these species pairs. We suggest that specialization to seasonally flooded habitats may explain patterns of adaptive radiation in many tropical tree genera and thereby provide a substantial contribution to regional tree diversity.     

Impact of the quality of climate models for modelling species occurrences in countries with poor climatic documentation: a case study from Bolivia

The quality of climate models has largely been overlooked as a possible source of uncertainty that may affect the outcomes of species distribution models, especially in the tropics, where comparatively few climatic stations are available. We compared the geographical discrepancies and potential conservation implications of using two different climate models (Saga and Worldclim) in combination with the species modelling approach Maxent in Bolivia. We estimated ranges of selected bird and fern species biogeographically restricted to either humid montane forest of the northern Bolivian Andes or seasonal dry tropical forests (in the Andes and southern lowlands). Saga and Worldclim predicted roughly similar climate patterns of temperature that were significantly correlated. Precipitation layers of both climate models were also roughly similar, but showed important differences. Species ranges estimated with Worldclim and Saga likewise produced different results. Ranges of species endemic to humid montane forests estimated with Saga had higher AUC (Area under the curve) values than those estimated with Worldclim, which for example predicted the occurrence of humid montane forest bird species near Lake Titicaca, an area that is clearly unsuitable for these species. Likewise, Worldclim overpredicted the occurrence of fern and bird species in the lowlands of the Chapare region and well south of the Andean Elbow, where more seasonal biomes occur. By contrast, Saga predictions were coherent with the known distribution of humid montane forests in the northern Bolivian Andes. Estimated ranges of species endemic to seasonal dry tropical forests predicted with Saga and Worldclim were not statistically different in most cases. However, detailed comparisons revealed that Saga was able to distinguish fragments of seasonal dry tropical forests in rain-shadow valleys of the northern Bolivian Andes, whereas Worldclim was not. These differences highlight the neglected influence of climate layers on modelling results and the importance of using the most accurate climate data available when modelling species distributions.